The present invention relates generally to the field of printed circuit boards, and, more particularly, to apparatus and methods for interconnecting stacked circuit boards and component layout on printed circuit boards.
Modern high performance computers contain a number of printed circuit boards (PCB) to carry and interconnect the various integrated circuit chips and other components that make up the computer system. Computers may be configured to have one or more subsystems, each in its own right a computer, that are interconnected into a larger, greater capacity computer. Each subsystem may contain basically the same components assembled in a chassis. Within a given subsystem, there may be, among others, a processor board, a power board, and any number of secondary or daughter boards that are carried upon a support structure within the subsystem chassis. These subsystems of boards are carried in racks in the computer chassis.
Processor boards are so named as they contain one or more processors, or central processing units (CPU) The CPU is an integrated circuit (IC) chip that is considered the xe2x80x9cbrainsxe2x80x9d of the computer. The processor board serves as a communication medium for the exchange of electrical signals among the one or more attached processors and other electrical components attached to the processor board. The processor board is sometimes referred to as a backplane, mother board, system board, or mainboard, depending on its function and configuration.
The processor board is a printed circuit board having one or more processors attached. A PCB is a relatively thin flat sheet structure. A PCB may be made of laminations of reinforced fiberglass or plastic and metallic interconnects which electrically link the components attached to the board. Since the interconnection level for the attached components is planar, the PCB is said to define an x-y interconnection plane.
The new generation of large processors take up considerable surface area of the processor board to which they are attached. In some cases, components which are typically attached to the PCB are moved to other locations on the board or moved entirely off the board onto daughter boards to make room for the processors. This disadvantage is compounded with the desire to put multiple processors onto one processor board. As a consequence, the distances between electrical components become large which is detrimental to the performance of the system as a whole.
Another electronic component attached to the processor board is an IC chip-set commonly known as the HUB. The HUB controls the computer""s electrical components such as, among others, the one or more processors, the memory, and the communication between other PCBs and external peripherals. In the recent past, processor board configurations provided for one processor controlled by one HUB. In order to increase the performance of a computer, designers have added additional processors on a single processor board creating multiple processor boards. A limitation to a computer""s performance, among others, is caused by the distance between electrical components, and in particular, the distance between the HUB and the one or more processors. As these distances become greater, the performance of the computer is compromised.
Computer memory chips are one of the critical components which are commonly attached to the processor board in order to be in close proximity to the HUB and the one or more processors. Close proximity is crucial as the time it takes for the HUB to communicate with the memory directly effects the performance of the system as a whole. This configuration is particularly crucial in very high speed supercomputer applications.
Due to the lack of physical surface area on the processor board after the attachment of the newer, larger processors, the memory chips must be moved to other locations on the board which increases the distance between the memory chips and the processors. Additionally, restricting the placement of the memory chips to the processor board would result in a limited number of chips that can physically be attached to a processor board of a certain size. This is unacceptable for computer systems requiring a large memory. As an alternative, the memory may be moved entirely off the board onto daughter cards. As a consequence, the distances between electrical components become large which is detrimental to the performance of the system as a whole.
Daughter cards are printed circuit boards which contain electronic components. Daughter cards are commonly configured to perform one of a number of specialized tasks, such as, among others, provide memory or display functions. Memory cards are daughter cards which contain a plurality of memory IC chips. Daughter cards are attached to the processor board in a number of ways. One popular method of attaching daughter cards to the processor board is by way of an edge connector. A daughter card configured to be used with an edge connector has electrical contacts along one edge of the PCB. The edge with the contacts is inserted into an edge connector socket attached to the processor board. Edge-connected daughter cards are commonly anchored along one edge by the edge connector itself and by a anchoring a second edge to the rack frame or chassis. Edge connectors generally require that the daughter cards be placed perpendicular to the processor board which adds to the overall dimensions of the computer. In an effort to make a smaller form factor computer chassis, this type of memory card configuration is unacceptable.
A high density interconnect configuration that has been utilized in the PCB art is a horizontal stacking of PCBs which are electrically interconnected with z-axis interboard connectors. This stacking and interconnection is referred to at z-axis interconnection. This three-dimensional packing scheme creates a compact module assembly that allows for a smaller chassis form factor.
Disadvantages with state of the art z-axis circuit board configurations is that unacceptably large interconnect distances remain between the daughter card and the processor, particularly for memory daughter cards. The physical limitation of placement of the z-axis connector, which attaches the daughter card to the processor board, with respect to the newer, larger processors, causes these large interconnect distances. Improved apparatus, methods and configurations are needed which provide for smaller interconnect distances between z-axis memory daughter cards and the processor.
Horizontally stacked daughter cards utilizing z-axis interboard connectors require an improved method and apparatus for daughter card support. In some cases, the daughter cards are not in proximity to the computer rack or chassis to which it could be anchored. Further, edge anchoring would not support the central portion of a large daughter card, or a daughter card having relatively heavy electronic components attached. Further, the z-axis interboard connector may not be suitable for carrying the stresses needed to support the daughter card.
Standoff mounts are used with varying success. Standoff mounts are devices which support and maintain a separation between two printed circuit boards or between the printed circuit board and the computer chassis. State of the art standoffs include those that are made for a chassis to PCB installation or are modifications of chassis-type standoffs used for interboard applications. Additionally, state of the art standoffs commonly contain many parts and are difficult to use in the confined spaces of the chassis of the computer. Part of the difficulty in using standoffs is alignment of the board mounting hole with the standoffs.
There exists a need for improved standoff apparatus, methods and configurations that are optimized for horizontally disposed, z-axis interconnected PCBs. Such an improved standoff will properly support the attached boards, facilitate alignment of the boards during assembly, not put undue stress on the boards, and provide for simple manufacturing.
For the reasons stated above, and for other reasons which will become apparent to those skilled in the art upon reading and understanding the present specification, there is a need in the art for improved methods and apparatus to address the disadvantages of large interconnect distances between z-axis mounted electrical components and the processor, and improved printed circuit board standoff methods and apparatus that are optimized for horizontally disposed, z-axis interconnected PCBs.
The above-mentioned disadvantages associated with large processors mounted on printed circuit boards and associated disadvantages of placing memory in close proximity to the processors, and other disadvantages, are addressed by the present invention and will be understood by reading and studying the following specification.
New apparatus, methods and configurations are provided that allows for a large memory capacity, as well as minimized interconnect distances between the memory chips and the one or more processors. The configuration and methods include placing the processors on one side of the processor board and the z-axis connector for the memory daughter cards placed on the opposite side of the processor board and in close proximity to the processor electrical leads. This enables the interconnection distance between the processor leads and memory daughter card leads to be minimized and not restricted by the physical size of the processor and z-axis connector housings.
In particular, an illustrative embodiment of the present invention includes a printed circuit board (PCB) having a plurality of apertures piercing through the PCB, wherein electrical leads of electrical components may pass through and be soldered. The PCB has a top surface and a bottom surface. The pattern of apertures on either side is substantially the same as the apertures pierce through the PCB. A plurality of groupings of apertures corresponding to the electrical lead patterns of larger electronic components are provided. Specifically, two HUB-chip conductive portions are provided, each HUB-chip conductive portion adapted to receive a HUB-chip. Additionally, four processor-chip conductive portions are provided. Each processor-chip conductive portion is adapted to receive a processor-chip. Two processor-chip conductive portions are disposed on opposite sides of each HUB-chip conductive portion in a symmetrical, mirror image arrangement. Eight z-axis connector conductive portions are provided, each z-axis connector conductive portion adapted to receive a z-axis connector. The z-axis connector conductive portion, the HUB conductive portion and the processor conductive portion are in close proximity to each other, minimizing the interconnect distance between the associated components.
Two HUB-chips, eight z-axis connectors, and various other electronic and mechanical components are attached to one side of the PCB. Four processor assemblies are attached to the converse side of the PCB. This configuration allows for the attachment of z-axis memory daughter cards to the z-axis connectors while retaining a minimum interconnect distance between the memory cards and the HUB.